Drift and/or Interference as Triggers of the Evolution

of Syntactical Patterns and Their Morphemic Markers:

The Case of the Evolution of Old Arabic into Neo-Arabic

Federico Corriente Saragossa University (Spain)

Linguistic change is the result of the replacement of either phonemic or morpho-syntactical patterns
in the structural levels of grammar or, in the realm of lexicon, the consequence of lexical substitution. In
any case, evolution, the inseparable companion of living languages, is most often triggered by internal
factors, which can be grouped together under the label of drift1 or, some other times, by external ones,
like the interference of a linguistic adstratum in the frequent case of languages in contact. It is generally
acknowledged that phonemic and lexical changes are very common in the latter case, as well as some
kinds of syntactical influences, while morphological borrowing happens very seldom, even between
closely related languages sharing one and the same area for quite long periods.
Consequently, linguistic change can be expected to have only internal causes most of the time, and
exceptions claimed to this principle must be carefully checked, especially when it is challenged too often,
as in the case of Comparative Semitics, where theories like that of Hebrew as a Mischsprache, Neo-Arabic
as the outcome of contamination of Old Arabic by foreign converts to Islam or, more recently, Amorite as
a melting pot of Semitic languages, have been put forward successively, in the manner of what is
commonly called a fad and far beyond what is expectable by strict and customary linguistic patterns.
Of course, it is well-known that, while Indo-Europeans spread since early times over large areas of
Europe and Asia and were more often than not in long-lasting and close contact with languages that were
not cognate of theirs, the Semites of old ages never overstepped the relatively confined lands of the Near
East in significant numbers, remained most of the time in communication with each others, ventured only
slowly and not too far away into the neighbouring countries of non-Semitic stock and developed two
remarkable panchronic socio-linguistic traits, which have accompanied them steadily through millennia,
namely, the easy acceptance of successive hegemony of one language of their group, from Akkadian to
Aramaic to Arabic, and the endeavour to achieve fast and total assimilation of included aliens as well as
their own kin.
Under such conditions, with large numbers of prospective bilingual speakers of two or more
cognate languages sharing the same country for centuries and even millennia, one must indeed expect
slower and lesser dialect differentiation, widespread imperfect bilingualism, pervasive code-switching, the
emergency of a Sprachbund situation and, in short, as much linguistic interference as can be allowed in
the realms of phonemic, syntax and lexicon, as is easily demonstrated by an even cursory examination of
Comparative Semitic grammars and dictionaries. Morphology, however, is known to withstand foreign
interference more stubbornly almost always in all situations of languages in contact; therefore, one must

1. In this paper we follow the definition of drift given for Neo-Arabic by Blau 1965: 12-18.

Aula Orientalis 26 (2008) 17-23 17

 FEDERICO CORRIENTE

scrutinize every particular alleged case of the kind between Semitic languages. As a sample case, we mean
to survey the rather abrupt evolution from Old Arabic to Neo-Arabic, traditionally attributed by both
native and foreign scholars to the disruptive agency of the muwallads or fast Arabicized inhabitants of the
Muslim Empire, who at first were by and large a majority of former speakers of Aramaic2. Incidentally
and not surprisingly, morphology is closely knitted into the syntactical categories meant by its markers,
which means that not only morphological features tend to exhibit a lower rate of replacement than in the
case of phonological traits and lexical items, but also that the syntactical patterns connected with those
synthetic features will have a higher life expectancy than phrase types merely using analytical markers,
such as word order or functionals.
Our method will be to survey the main morpho-syntactical features differentiating Old Arabic from
Neo-Arabic, according to Fischer & Jastrow (1980: 39-48, 61, 70-74 and 97-100), i.e., noun declension,
dual and mood marking, congruence, numerals and verbal measures and voices, looking for the most
likely reason of the corresponding shifts in those realms, whether triggered by drift or interference of the
linguistic substratum, in order to gauge the relative impact of both factors in the evolution of the best
known Semitic tongue.
1) Noun declension: as it was abandoned by every Semitic tongue which lived long enough to see
the total forsaking of a low-yielding synthetic device with only three cases at most3, marked by short
vocalic suffixes, one could simply assume that this was also the case of Arabic in which, besides, the early
and regular decay of final vowels in pause positions further decreased that scarce functional yield4.
However, the fact that some Bedouins kept the Old Arabic noun declension system alive until the 10th
century5, while the bilingual populations of Syria and Iraq had lost it before the rise of Islam, suggests that

2. Perhaps the best extant account of scholarly positions on this thorny issue is Blau 1977, where he sticks to his hypothesis of
a post-Islamic emergency of the Neo-Arabic type, but allows the possibility of the emergence of the Neoarabic linguistic type
before Islam with tribes that did not partake in the culture of Standard Arabic poetry, which is, in our view, an appropriate
synthesis putting an end to the polite debate between both of us, in which the traditional ungranted thesis was confronted with a
partial antithesis (Corriente 1971 and 1973-4). For there is no point in denying that the central core of Arab tribes which produced
and enjoyed old poetry must have had Old Arabic as their native tongue, and that the lands in which Qurn was first preached
also partook of the same situation, even at the time when Neo-Arabic was surfacing to the North and South of those areas and
only waiting for its chance to win the demographic battle in the aftermath of the upheaval caused by the Islamic expansion.
3. One could be led into confusion on this point by Diakonoff 1965: 58, where Old Akkadian is said to have preserved the
most complete inventory of Proto-Semitic cases, with five items. In fact, the absolute form, marked with or {} or {-a} is
phonetically, not morpho-syntactically, conditioned, while the author himself acknowledges that it is very difficult to say
anything definite of the Dative-Locative case. Short case inventories are inefficient and must be supplemented with analytical
devices which, in turn, make case marking irrelevant and dispensable (on this, see Corriente 1971: 44-50). But except when an
adstratum is clearly at work (as in the case of Bulgarian, the only Slavonic language to have lost case marks), it is not easy to
pinpoint the cause of the mergers or foretell the time when drift will begin acting and lopping off a more or less rich case
inventory until its final and complete elimination: for a five or six case system it did happen soon enough in Latin, but only
partially in Modern Greek until this day, and not at all, at least yet, in Slavonic, with the aforementioned exception. The drastic
reduction to two and even total elimination of the case feature is also found in the four-case old Germanic declension. But one
thing appears to be sure, namely, that very short inventories, i.e, two or three cases only, do not often hold on too long, like in the
cases of Semitic languages and Old French, while long inventories, like those of Finnish and Turkish, much more efficient and
less dependant on auxiliary functionals, are more impervious to coalescence and reduction.
4. The role of pausal forms in the decay of Arabic noun inflection has always being considered crucial. By the mere principle
of least effort, when each of the members of a paradigm can be used with or without a given additional suffix, this is analysed as
devoid of any function, a mere phonetic appendix and is finally dispensed with, so that the whole paradigm disappears from the
morphemic inventory of the language.
5. See Corriente (1986: 66, fn. 3). As we say in a forthcoming paper, the famous papyrus unearthed by Violet 1901 in the
Umayyad Mosque of Damascus, dated from the end of the 8th century and already drawn up in Neo-Arabic but for a few residual
Old Arabic features, may be considered the birth certificate of Neo-Arabic, although issued a couple of centuries after the birth
date.

18
 DRIFT AND/OR INTERFERENCE AS TRIGGERS OF THE EVOLUTION OF SYNTACTICAL PATTERNS

Aramaic interference may have accelerated this process among the so-called anb, i.e., Arabs who had
settled in those areas some centuries before Islam. Most likely the same situation may be posited also for
the Arabs of Southern origin, wherever their abode was, on account of the doubts existing about the
vitality of noun declension in Epigraphic South Arabian6, underscored by its total disappearance in
Modern South Arabian7 and its peculiar status in G#c#z8. In other words, drift would have been sufficient
to bring about the loss of noun declension in the evolution from Old Arabic to Neo-Arabic, but it appears
reasonable to also admit the corroborating effect of interference in the same areas that would be of
paramount importance in the emergency and consolidation of the Neo-Arabic type9.
2) Dual marking: by forsaking the category of dual in verbs, pronouns and adjectives and allowing
it only in more or less restricted classes of substantives, Neo-Arabic has reverted from the hypertrophic
situation of Old Arabic to the standards of Semitic. As morphological hypertrophy is a hallmark of South
Semitic10, in contrast with its Eastern and Northern branches, chances are that Neo-Arabic has been
influenced in this matter by the Aramaic substratum of the bilingual anb, but again only as a
corroborating factor, since dual marking is reputedly a primitive feature in regression everywhere.
3) Mood marking: Old Arabic had a seemingly well developed system of four moods (indicative,
subjunctive, jussive and energic)11 in the imperfective, marked by attached vocalic morphemes and the
alternance of presence, absence and gemination of /n/, while the rest of Southern Semitic exhibits a mere
opposition between indicative and subjunctive (also a jussive), marked by stem alternance, in ways
morphologically parallel to the Accadian present and preterite12. As the mood ending system of Old
Arabic is vaguely matched only by Ugaritic13 alone among all Semitic languages, our impression is that
Old Arabic had once again developed the morphosyntactic category of mood in its characteristic
hypertrophic manner. The disappearance of these marks in Neo-Arabic can be attributed to their loss of

6. See Brockelmann (1908 I: 463). Bauer (1966: 61) vouchsafes its existence only for the oldest periods and, for younger
Himiaritic, Belova (1996: 103-111) posits the reduction of the system to a minimal expression. Of course, we must agree with
those who, from Goldziher to Behnstedt, have advised caution in wholesale accepting the native reports on the ethnical division of
the Arabs, but solid linguistic evidence confirms its basic truth, with due allowance to exaggerations, mistakes and fabrications.
7. Jahn (1905: 69-70) underscores this fact for Mehri, while other researchers, like Johnstone 1975 do not even mention the
case feature in their monographs.
8. Where it has been cut down to a two-case system of accusative and non-accusative, characteristic of the penultimate stage
of many processes of loss of noun declension, which has been completed in Modern Ethiopic languages, but for marginal
phenomena such as the reintroduction of an accusative mark in Amharic: see Cohen (1970: 81).
9. By saying this we do not have in mind just the impact of the anb in the emergency of Neo-Arabic, but also that of former
speakers of South Arabian (the ubiquitous Yemenites) in the differentiation of Western Arabic dialects, as outlined in some of
our writings since Corriente 1989. We are aware of the strictures made to this hypothesis by our good friend and colleague
Behnstedt 2004, who offers alternative explanations to the phenomena surveyed by us, but only to some of them, as he
acknowledges in p. 342 that his purpose is einige dieser bereinstimmungen ablehen bzw. relativieren. Which, of course, may
be in place and, coming from an expert like him, must be thanked for, as any other effort to improve our present state of
knowledge of the beginnings of Arabic dialectology.
10. As witnessed to by Modern South Arabian, where dual has been introduced even in the first person of verbs (see
Johnstone 1975: 15) which, as the author reminds us, has only heretofore been recorded for Ugaritic.
11. However, as the selection of mood is often conditioned by the appearance in the context of some functionals or syntactical
sequences, and not merely by the corresponding logical categories, their true functional yield is rather low, it being frequent to
have indicative for a clearly subjunctive nuance (e.g., m raaytuh yaktubu I did not think that he would write), or the other
way around (e.g., lan yaktuba he will not write); the same happens with the opposition between indicative and jussive moods
(e.g., yajlisu cAmrun wayanarifu bqi lmu\annn let cA. sit down and the other singers go away, from Corriente [1975: 51],
where the vocalization must be corrected as done here, vs. lam yaktub he did not write).
12. Which has been often claimed as proof of a particularly close kinship between Akkadian and peripheral South Semitic
(excluding Arabic), in spite of Rundgrens warnings and precisions about the exact reach of that outer similarity (see also
Corriente [2003: 191]).
13. See Gordon (1965: 71-73).

19
 FEDERICO CORRIENTE

any residual functional yield in pausal forms, where short final vowels were necessarily dropped, as well
as to the cooperation of the Aramaic substratum14. However, most dialects could not do without at least a
minimal opposition between subjunctive and indicative imperfective, the latter being often marked
through analytical means, such as particles, verbs, or what remains of them through tear and wear upon
becoming particles15, a system not alien to Aramaic either.
4) Congruence: Old Arabic practised an odd kind of congruence, which is not totally what
linguistics call natural. i.e., solely hinging upon the concepts of gender and number. Instead, it still
partially reflects a primitive stage of Afrasian languages, in which noun classes overlapped those two
concepts16 in intricate ways not altogether disentangled by contemporary linguists, in spite of enlightening
comparison with the situation in Hamitic and Bantu languages. In any case, that is the reason beyond the
peculiar agreement in Old Arabic between irrational plural nouns with feminine singular (more seldom
plural) adjectives and verbs (e.g., camal-un li a pious deed but acml-un li-at-un pious deeds,
also li-t-un, while rational plural nouns, whether masculine or feminine, agree with plural masculine
or feminine adjectives and verbs in what we call a natural way (e.g., muslim-un liun a pious Muslim,
muslim-na li-na pious Muslims, muslim-at-un li-at-un a pious Muslim woman, muslim-t-un
li-t-un pious Muslim women)17. Native grammarians tell us that many Arabs only knew the totally
natural agreement familiar to us, called by them lu\atu akalni lbar\,18 as reported by the father of
Arabic grammar, Alxall b. Amad, in all types of Old Arabic, including poetry and Qurn, which put
purists to the difficult ask of explaining this anomaly19. As for Neo-Arabic, and in the lack of a detailed
comparative survey of this matter taking into consideration every dialect or at least a large enough sample
of them, it could be said that natural agreement is the rule, with frequent exceptions due to survival or the
interference of the old or high registers, in a wide range from zero, in the extreme case of Maltese, early
rid from that interference, up to nearly optional agreement in feminine singular of irrational plurals in both
Eastern and Western dialects20. All in all, though not thoroughly rejecting the possible effect of

14. This, however, cannot have affected the merger of moods in persons marked by the presence or absence of a final /n/, i.e.,
2nd feminine singular, duals and masculine plurals, as that consonant is always extant in Aramaic in those persons, while Neo-
Arabic dialects exhibit a cleavage on this point, after the loss of mood marking, between two groups neatly differentiated by total
phonetic preservation of that former mark (Eastern Arabia and Iraq) or its complete loss (all the others, according to Fischer &
Jastrow [1980: 42]). In this instance, as in that of the survival in every younger phase of Semitic languages of the oblique case in
duals and masculine plurals, the explanation cannot lie in the interference of the substrata, but in drift and arithmetical option,
since oblique cases and moods are statistically more frequent that its counterparts and therefore likelier to win the upper hand in
the fight for survival of indifferent doublets.
15. See Fischer & Jastrow (1980: 74-75).
16. See Brockelmann 1901 I: 404-405, Fleisch (1961: 312-338), Diakonoff (1965: 55-56) and Diakonoff (1988: 57-59).
17. Another striking congruence rule in Old Arabic, to be forsaken in Neo-Arabic, is the compulsory singular of verbs
preceding their subjects, whether singular, dual or plural, which is a phenomenon of an entirely different nature, namely, a case of
marking economy, comparable to the use of the singular with numbered items over ten, also found in Persian and Turkish with
any numeral over one. Incidentally, the very restricted use of plural in numbered items from 3 to 10, coinciding with the so-called
pluralis paucitatis, is again a remnant of a noun class system opposing countable = important vs. non-countable = non-important
nouns, and so are the singulatives obtained with the feminine mark, the femine plural marking of diminutives, etc.
18. I.e., the dialect of those who say akalni lbar\, instead of akalatn for the fleas ate me, using the plural of the verb,
not the 3rd person feminine singular, as demanded by the Classical Arabic standards. By the way, that shibboleth would have been
completed by saying akalni lbar\u lkibr big fleas ate me, instead of the standard akalatni lbar\u lkabrah.
19. See Corriente (1976: 75).
20. This would also be the case of the oldest Neo-Arabic dialect for which we have extensive records, i.e., Andalusi,
according to Corriente (1977: 130), but the optionality appears to be general: cf. , for the Egyptian dialect, Mitchell (1956: 24, fn.
1), the adjective accompanying plural nouns other than those referring to human beings - is almost always in the feminine
singular, Baskinta (Lebanon), Abu-Haidar (1979: 124), With plural substantives which do not refer to human beings the
qualifying adjective is always feminine singular where the substantive is abstract, Where the substantive is concrete the

20
 DRIFT AND/OR INTERFERENCE AS TRIGGERS OF THE EVOLUTION OF SYNTACTICAL PATTERNS

interference on this particular issue, the strong shift towards natural agreement in Neo-Arabic appears to
be mostly a matter of drift.
5) Numerals: the reassignment of the former shapes of the numerals from 3 to 10, marked for the
two genders, to a new distribution, depending on whether or not the numbered item follows, has been
pointed out as another hallmark of Neo-Arabic, although some dialects exhibit further deviations from that
new distribution. The fact that this development has not taken place in the dialects of the Arabian
Peninsula could give strength to the hypothesis of interference by a foreign substratum, which, however,
cannot be Aramaic, being as it is akin on this point to Old Arabic21. In all likelihood only drift caused this
phenomenon in a very weak point of the morphemic structure of Semitic grammar, in which the gender
marks were used in a manner counter to their usual function, i.e., {-at} for the masculine and {} for the
feminine. Linguistic usage chose to keep the old enumerating series with the feminine mark, while
analogical evolution eliminated the anomalous gender distinction by coalescence into the unmarked form
for numerals followed by numbered items.
6) Verbal measures and voices: the forsaking in Neo-Arabic of the least frequent old derived verbal
patterns or measures22 and of the internally marked non-agentive voice has long since been underscored,
as well as the loss of the IV measure, except in the latter case for Eastern Bedouin dialects23. In this
instance, the usual explanation for its decay has been the weakness of the phoneme // as a mark which
already in Old Arabic disappeared in the imperfectives24, which may be true, but only a part of the truth.
As a matter of fact, the same kind of weak markedness, above all in the imperfectives, has not prevented
the causative derived verbs from remaining alive and well in the Aramaic apcel and even thriving in
Southern Ethiopic25. In our view, the very instability of // turned the vowel opposition into the most
reliable mark of this category, so that when an evolution of the vowel system happened at some point in

qualifying adjective is usually plural , Christian Baghdadi, Abu-Haidar (1991: 102), Broken plurals substantives, referring to
animals and inanimate objects are frequently qualified by feminine singular adjectives, Gulf Arabic, Qafisheh (1977: 234),
Adjectives modifying non-human plural nouns are usually feminine singular, but may be plural; the latter form is not commonly
used, Oman, Reinhardt 1894: 56-57, Dem Gebrauch nach sind weiblich d) Alle gebrochenen Plurale mit Ausnahme der
lebende menschliche mnnliche Wesen bezeichnenden Haupwrter, Eastern Sudan, Reichmuth 1983: 200, Typ 1.
Zhlkongruenz: bei adjektiven Plural Stmmen, soweit vorhanden, ansonsten Pl. m. fr menschliche Maskulina, Pl. f. in allen
anderen Fllen Typ 2. Mengenkongruenz: Sg. f. in allen Fllen, genusunabhngig Ansonsten kommen Typ 1 und Typ 2
nebeneinander vor.. , Moroccan, Buret (1944: 80-81), Vous trouverez souvent, dans les texts, des verbes, ou des adjectives, au
feminine singulier se rapportant des noms au pluriels. Dans certains cas cela tait obligatoire en arabe classique. Aujourd hui,
dans certaines expressions, on emploie volontier le feminine singulier, mais il n est pas incorrect d employer le pluriel, the
Algerian dialect, Tapiro (1971: 42, fn. 1), Un pluriel d objects ou d animaux entrane souvent un accord au f. sg.,
Maghribian as a whole, Marais (1977: 158), la tendence dialectale gnrale consiste faire l accord au pluriel des verbes
(adjectives et pronoms) qui se rapportent des noms comportant l ide manifeste de pluralit Il est cependant courant, surtout
dans les parlers bdouins et dans les parlers de l Est maghrbin, de relever l accord au feminine des verbes (adjectives et
pronoms) se rapportant des pluriels et collectifs, etc. Summarizing, as stated by Fischer & Jastrow (1980: 96), Zum Plural von
Sachbezeichnungen wird das Adjektiv in vielen Dialekten in femininen Singular gestellt. Jedoch gibt es andere Dialekte, in
welchen auch in solchen Fllen der Plural des Adjektivs blich ist. Nicht selten sind beide Mglichkeiten gegeben.
21. Lack of gender distinction in these numerals is characteristic of Greek and Persian, but neither of them was in a position as
a substratum strong enough to force such a shift in Arabic grammar. The same can be said of Coptic which, however, has also a
situation similar to that of Old Arabic in what concerns gender distinction of numerals.
22. I.e., from XII to XV, and IX-XI in many areas, for triconsonantic, and III and IV for quadriconsonantic roots. Their
abandonment is a clear consequence of the trend towards reversion from hypertrophy characteristic of Neoarabic, as in the case
of proliferation of the broken plural and madar patterns, etc.
23. See Fischer & Jastrow (1980: 46).
24. Cf. Fischer & Jastrow (1980: 46): Die erwhnte Schwche des hat in den meisten Dialekten zum Verlust des IV.
Stamms afcala gefhrt. Brockelmann (1901 I: 523) felt likewise when he wrote: In den marib. Dialekten wird das
Verschwinden des IV. Stammes noch durch den lautlichen Verlust des Kausativprfixes begnstigt.
25. E.g., in Amharic adkkm, annaggr, etc.: see Cohen (1950: 222-226 and table XXII).

21
 FEDERICO CORRIENTE

Old Arabic the momentous distinction in the imperfective of many couples of verbs, intransitive I
measures (e.g., yaxruj he goes out) vs. transitive IV measures (e.g., yaxrij he takes out) became
threatened. Fischer & Jastrow (1980: 43-44) have signalled the weakening of the opposition between /i/
and /u/ in Neo-Arabic and pointedly suggested it as the reason for the disappearance of the internally
marked non-agentive voice26; they could as well have extended this reasoning to the merger of I and IV
measures.
Summing up: while foreign interference indeed appears to have acted as a corroborating factor,
occasionally perhaps even a strong factor, in the evolution from Old Arabic to Neo-Arabic, drift has been
most of the time the initial or even sometimes only motive of the most conspicuous changes that have
taken place between both phases of this language. This said, our statement would not be complete without
adding that, in our view, the two main exceptions to the overwhelming power of drift in those changes, are
the important role of the anb in the emergence and consolidation of the common core of features of most
modern Arabic dialects, and that of former speakers of South Arabian in the development of Western
Arabic dialects.

BIBLIOGRAPHY

Abu-Haidar, F. 1979. A study of the spoken Arabic of Baskinta, LeidenLondon.

Abu-Haidar, F. 1991. Christian Arabic of Baghdad, Wiesbaden.
Bauer, G. 1966. Jazyk yunoaravijskoj pismennosti, Moscow.
Behnstedt, P. 2004. Bezge zwischen maghrebinischen und jemenitischen Dialekten, Jerusalem Studies
in Arabic and Islam 29:341-357.
Belova, A.G. 1996. Ximjaritskij jazyk. Arealnye issledovanija k istorii arabskogo jazyka, Moscow.
Belova, A.G. 1999. Oerki po istorii arabskogo yazyka, Moscow.
Blau, J. 1965. The emergence and linguistic background of Judaeo-Arabic. A study of the origins of
Middle Arabic, Oxford.
Blau, J. 1977. The beginnings of the Arabic Diglossia: A Study of the Origins of Neo-Arabic,
Afroasiatic Linguistics 4/41:1-28, rep. in Studies in Middle Arabic and its Judaeo-Arabic variety,
Jerusalem, 1988:1-38.
Brockelmann, C. 1908. Grundri der vergleichenden Grammatik der semitischen Sprachen, Berlin (rep.
Hildesheim, 1966).
Buret, M.-T. 1944. Cours gradu d arabe marocain, Casablanca.
Cohen, M. 1970. Trait de langue amharique (Abyssinie), Paris.
Corriente, F. 1971. On the functional yield of some synthetic devices in Arabic and Semitic
morphology, Jewish Quarterly Review 62:20-50.
Corriente, F. 1973-4. Again on the functional yield of some synthetic devices in Arabic and Semitic
morphology, Jewish Quarterly Review 64:154-163.
Corriente, F. 1975. Marginalia on Arabic diglossia and evidence thereof in the Kitb al-a\n, Journal
of Semitic Studies 20:38-61.

26. But the occasional lack of differentiation of the closed vowels in Semitic is much older, of course, and obeys to an array
of causes (see Belova 1999: 36-40). Even leaving aside the possibility of eventual preservation of a Proto-Semitic two-vowel
system (suggested by Diakonoff, Militarev, Porkhomovskiy and others), certain minor alterations of phonemic elements, like
increasing or decreasing strength of the tendency of vowel harmony, or changes in the nature of emphasis from glottalization
towards velarization or pharyngealization, which immediately becomes a suprasegmental affecting the whole lexical unit, may
play havoc with a rather poor vowel inventory and destroy morphological oppositions merely based upon it.

22
 DRIFT AND/OR INTERFERENCE AS TRIGGERS OF THE EVOLUTION OF SYNTACTICAL PATTERNS

Corriente, F. 1976. From Old Arabic to Classical Arabic through the pre-Islamic koine: Some notes on
the native grammarians sources, attitudes and goals, Journal of Semitic Studies 21:64-98.
Corriente, F. 1977. A grammatical sketch of the Spanish Arabic dialect bundle, Madrid.
Corriente, F. 2003. On the degree of kinship between Arabic and Northwest Semitic, AIDA 5th
Conference Proceedings, Cdiz.
Diakonoff, I.M. 1965. Semito-Hamitic languages. An essay in classification, Moscow.
Diakonoff, I.M. 1988: Afrasian Languages, Moscow.
Fischer, W. & Jastrow, O. 1980. Handbuch der arabischen Dialekte, Wiesbaden.
Fleisch, H. 1961. Trait de philologie arabe I, Beirut.
Gordon, C. 1965. Ugaritic textbook, Rome.
Jahn, A. 1905. Grammatik der Mehri-Sprachen in Sdarabien (Sitzungsberichte der philosophisch-
historischen Klasse der kaiserlichen Akademie der Wissenschaften, 150), Viena.
Johnstone, T.M. 1975. The Modern South Arabian Languages (Monographic Journals of the Near East),
Malibu CA.
Marais, Ph. 1977. Esquisse grammatical de l arabe maghrbin, Paris.
Mitchell, T.F. 1956. An introduction to Egyptian colloquial Arabic, Oxford.
Qafisheh, H.A. 1977. A Short Reference Grammar of Gulf Arabic, Tucson AZ.
Reichmuth, S. 1983 Der arabische Dialekt der ukriyya im Ostsudan, Hildesheim-Zrich-New York.
Reinhardt, C. 1894. Ein arabischer Dialekt gesprochen in cOman und Zanzibar, Stuttgart (rep.
Amsterdam, 1972).
Rundgren, F. 1959. Intensiv und Aspekt-Korrelation. Studien zur thiopischen und akkadischen
Verbalsammbildung, Uppsala-Wiesbaden.
Tapiro, N. 1971. Manuel d arabe algrien moderne, Paris.
Violet, B. 1901. Ein zweisprachiges Psalmfragment aus Damaskus, Orientalistisches Litteratur-Zeitung
4, 384-403, 425-441, 475-488.

23